A Hybrid Method of Feature Subset Selection
スポンサーリンク
概要
- 論文の詳細を見る
Feature subset selection is of prime important in pattern classification, machine learning and data mining applications. A real world database may contain many noisy, unnecessary and irrelevant features. If it is used for data minihg directly, the quality of the discovered knowledge may be very poor. To cope with this problem, many methods have been proposed. In this paper, we propose a hybrid algorithm by using class mutual information for feature selection, starting from the Rough Sets CORE. If the CORE is empty we use binary mutual information for the first feature selection. Experiments have been conducted on some artificial and real world domains in terms of tree size, test errors rate and subset sizes. The results show the effectiveness of proposed hybrid algorithm.
- 日本知能情報ファジィ学会の論文
- 2002-12-15
著者
-
ISHII Naohiro
Department of Intelligence and Computer Science, Nagoya Institute of Technology
-
Ishii Naohiro
Department Of Intelligence And Computer Science Nagoya Institute Of Technology
-
Ishii Naohiro
The Faculty Of Engineering Nagoya Institute Of Technology
-
BAO Yongguang
Aichi Information System
-
DU Xiaoyong
School of Information, Renmin University of China
-
BAO Yongguang
Department of Intelligence and Computer Science, Nagoya Institute of Technology
-
Du Xiaoyong
School Of Information Renmin University Of China
-
Bao Yongguang
Department Of Intelligence And Computer Science Nagoya Institute Of Technology
-
Ishii Naohiro
Department Of Electrical Engineering And Computer Science Nagoya Institute Of Technology
関連論文
- Neural Network Based Photometric Stereo with a Nearby Rotational Moving Light Source
- Moving Point Light Source Photometric Stereo
- Photometric Stereo for Specular Surface Shape Based on Neural Network (Special Issue on Neurocomputing)
- Shape Recovery of Glossy Surface from Shading Image and its Directional Derivatives
- The Completeness of Order-Sorted Term Rewriting Systems Is Preserved by Currying
- An Energy-Efficient Initialization Protocol for Wireless Sensor Networks with No Collision Detection
- Termination of Order-Sorted Rewriting with Non-minimal Signatures
- Accuracy of the Minimum Time Estimate for Programs on Heterogeneous Machines
- A Lookahead Heuristic for Heterogeneous Multiprocessor Scheduling with Communication Costs (Special Issue on Parallel and Distributed Supercomputing)
- An Improved Increase over the Minimum Execution Time of a Parallel Program
- An Efficient Method for Computing All Reducts
- A Hybrid Method of Feature Subset Selection
- Improving Performance of the k-Nearest Neighbor Classifier by Combining Feature Selection with Feature Weighting
- Speeding up String Searching Algorithms for Nonuniform Texts
- On Relationships between Decomposable Programs and Rule Commutative Programs
- Efficient Evaluation of One-directional Cycle-recursive Formulas
- A Performance Measure for the Scheduling of Typed Task Systems with Communication Costs
- Decomposable Programs Revised
- Determining Feature Weight of Pattern Classification by Using Rough Genetic Algorithm and Fuzzy Similarity Measure
- Fast On-line String Searching
- Reliability of a Mobile Communication System with Network Congestion
- Optimizing Linear Recursive Formulas by Detaching Isolated Variables